[BioC] positively correlated genes

Gordon K Smyth smyth at wehi.EDU.AU
Sat Sep 13 01:10:18 CEST 2014


If the logFC is positive, the gene is positively correlated with gene X.

If the logFC is negative, the gene is negatively correlated with gene X.

You should be using the FDR column rather than the PValue column to judge 
significance, as is usual with edgeR analyses.

Pearson correlation between y and x is equivalent to regressing y on x. 
Pearson correlation is significant and positive if and only if the 
regression coefficient of y on x is significantly positive.  Situation is 
similar here.

Gordon


On Fri, 12 Sep 2014, Sindre Lee wrote:

> Can I please ask how to interpret this results? Im used to 
> Spearman/Pearson correlations and don't quite know how to present or 
> explain the results obtained this way.

I wanted to find genes correlating with gene X. Then I got about 6000 
significant genes at p < 0.05. Some with negative some with positive 
log2FC.

Now, what do I do? What does this tell me?


Thank you!


________________________________________
From: bioconductor-bounces at r-project.org <bioconductor-bounces at r-project.org> on behalf of Gordon K Smyth <smyth at wehi.edu.au>
Sent: 10 September 2014 03:08
To: deepaksrna at gmail.com
Cc: Bioconductor mailing list
Subject: [BioC] positively correlated genes

If you are using edgeR's glmFit function or limma's voom and lmFit
functions, you can simply add the log-expression values of the gene of
interest as a column of the design matrix.  Then a standard DE analysis
will detect any other genes that are significantly correlated with the
gene of interest.

Gordon

> Date: Tue,  9 Sep 2014 01:41:14 -0700 (PDT)
> From: "karthik [guest]" <guest at bioconductor.org>
> To: bioconductor at r-project.org, deepaksrna at gmail.com
> Subject: [BioC] positively correlated genes
>
> hi
>   I am interested to find out the genes that are positively and
> negatively correlated genes with my genes of interest. (using rnaseq
> normalized expression data). Can some one suggest me a better option.
>
> Thank you
>
> -- output of sessionInfo():
>
> sessionInfo()
> R version 3.0.2 (2013-09-25)
> Platform: x86_64-w64-mingw32/x64 (64-bit)

______________________________________________________________________
The information in this email is confidential and intend...{{dropped:4}}



More information about the Bioconductor mailing list